Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Science Project Help

eddieiseddie

eddieiseddie

macrumors member
Original poster
Jul 9, 2008
34
0
I'm working on a science project for school. Here's what I'm trying to do:

I am trying to figure out what affects a computer's (in my case my black MacBook 2.2)speed most, CPU, RAM, cache,etc. All i have right now is my MacBook to test on, it has 4gig ram which I have upgraded from 1gig. Now, is it ok to mix the original ram that came w/ the mac with the new ram to make laptop have 2.5gigs of ram? Also i also have a cousin w/ a black MacBook too:), but the gen before mine, i think it's 2.0 ghz, that i might be able to test on. The comps at my school are all dells, so idk if it would be a good idea to mix the platform of os during the experiment; but i could also test which os runs faster while loading certain webpages, or something like that:confused:...So you can easily see how many variable I have to consider in this project:rolleyes:

Have you cool cats got any suggestions on how I should start getting some good data? What makes it even worse is that it's due by this Tuesday!:eek:
Ya, I have a week from today to do this, so please keep that in mind.

Thanks so much in advance!:D
 
Yeah, you can mix and match memory to your heart's content. I'm not sure Macs are the machines to be doing this testing on, though, particularly Mac laptops. Memory is the only component that you can easily swap out. I'd suggest finding an old PC you can mess with to your heart's content.

For an idea of variables and methods of testing, take a look at some benchmark sites and see what they pay attention to.
 
You're best bet is a pc desktop, with 3 different videocards / cpus / sets of memory. Using the mid level one of each of those as the control. Then swap them out in every order you can. Benchmarking as you go. It's not the easiest or cheapest thing to do, but it's you would get the best results.


Is this project of your picking?
 
too many variables that will be unexplained. too much variety in the hardware that can be swapped out. all 512mb video cards arent equal, because of differences in the technology behind it, all 1ghz CPUs arent euqal for teh same reason and all 1gb sticks of ram arent equal either. depends on many other components as well, like the MB, FSB and how they interact with the rest of your hardware and teh cpu, and how all of that interacts with your software. not to mention that this would cost you a fortune to set up

also depends on the type of application you are running. processor intensive, graphics intensive.

not to mention a science experiment should have controls. what could your controls possibly be?

i think a better project would be something along the lines of this (it would also be much more straightforward and you can better show knowledge of proper scientific technique), and if you could report your results, id love to hear it because ive always believed in this theory even though im not 100% sure if its true

the theory (and your hypothesis) would be that the FSB speed of your memory should be the same as that of your MB and CPU. if its higher, it has to be forced lower, which in turn diminshes system performance (because it will only operate at peak performance if its running at its native FSB). if its lower, the rest of your system has to compensate for the lower FSB RAM. Your null hypothesis would be the opposite (that syncing of the memory FSB with the rest of the system doesnt have anything to do with system performance), so your experiments would go like this:

your negative control would be a system with no RAM, doesnt turn on - cant measure

positive control would be memory that has a FSB equivalent to your MB and CPU

then your experimental group would be the same amount of memory (1gb, 512 mb whatever) but that has a FSB faster than your CPU and MB or slower than your CPU and MB, if your null-hypothesis is incorrect (and thus your hypothesis is correct), you would expect system performance to be diminished in either of the cases, and you can state that syncing the FSB of your memory with the rest of the system increaes the system's overall performance.

run some sort of system performance test, report results, get A+

*PS dont forget your replicates. run the test numerous times, resetting the computer between each one, average the results, and report the probable error values and if the results are significant in your data (off the top of my head i dont know what kind of stats you would have to do in order to get the error values or the ones in order to test for significance, if someone could fill that in for me).

ideally you would want to confirm your results by buying 2 or 3 sticks of each type of RAM, testing them independantly, but i think that may be going overboard in terms of scope of what you actually need to do.
 
wow Davy u really have an awesome layout for me to follow in doing this project:cool: thanks alot for ur help! if i end up going ur route ill make sure to let u know:D
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back